1 Setup

2 Tumor analysis

2.1 Shared mapping & helpers

group_name_map <- c(
  `11636.1` = "Control",
  `11636.2` = "Tamoxifen",
  `11636.3` = "Dexamethasone",
  `11636.4` = "Tamoxifen + Dexamethasone"
)

pal <- c(
  "Control"                   = "#1f77b4",
  "Tamoxifen"                 = "#ff7f0e",
  "Dexamethasone"             = "#2ca02c",
  "Tamoxifen + Dexamethasone" = "#d62728"
)

bin_day_keep0 <- function(day, width = 3L, anchor = 1L, label = c("start","center")) {
  label <- match.arg(label)
  base <- anchor + width * floor((day - anchor)/width)
  base <- ifelse(day <= 0, 0L, base)
  if (label == "center") {
    base <- ifelse(day <= 0, 0L, base + floor((width - 1L)/2L))
  }
  as.integer(base)
}

2.2 Tumor: data loading

mouse_df <- readxl::read_excel(tumor_file, sheet = "Mouse_day_mean") %>%
  dplyr::mutate(
    group_code = as.character(group_code),
    group_name = factor(unname(group_name_map[group_code]), levels = names(pal)),
    day        = as.numeric(day),
    mouse_id   = as.character(mouse_id),
    volume_mm3 = as.numeric(mean_volume_mm3)
  ) %>%
  dplyr::filter(group_code %in% names(group_name_map)) %>%
  dplyr::arrange(group_name, mouse_id, day)

knitr::kable(head(mouse_df, 8), caption="Preview: tumor per mouse per day")

Preview: tumor per mouse per day
group_code	mouse_id	day	mean_volume_mm3	group_name	volume_mm3
11636.1	2539021	0	34.5	Control	34.5
11636.1	2539021	2	28.5	Control	28.5
11636.1	2539021	4	38.0	Control	38.0
11636.1	2539021	8	31.5	Control	31.5
11636.1	2539023	0	50.5	Control	50.5
11636.1	2539023	2	49.0	Control	49.0
11636.1	2539023	4	61.5	Control	61.5
11636.1	2539023	8	101.0	Control	101.0

2.3 Tumor: data quality check

min_points_required  <- 4

tum_qc <- mouse_df %>%
  group_by(group_code, group_name, mouse_id) %>%
  arrange(day, .by_group = TRUE) %>%
  summarise(
    n_points   = n(),
    min_day    = min(day, na.rm = TRUE),
    max_day    = max(day, na.rm = TRUE),
    first_vol  = first(volume_mm3),
    last_vol   = last(volume_mm3),
    min_vol    = min(volume_mm3, na.rm = TRUE),
    max_vol    = max(volume_mm3, na.rm = TRUE),
    .groups    = "drop"
  ) %>%
  mutate(
    recommend_exclude = (n_points < min_points_required) | (min_vol <= 0)
  ) %>%
  arrange(group_code, mouse_id)

knitr::kable(
  tum_qc,
  caption = "Simplified tumor QC table per mouse"
)

Simplified tumor QC table per mouse
group_code	group_name	mouse_id	n_points	max_day	first_vol	last_vol	min_vol	max_vol	recommend_exclude
11636.1	Control	2539021	4	8	34.5	31.5	28.5	38.0	FALSE
11636.1	Control	2539023	26	61	50.5	102.0	46.5	310.0	FALSE
11636.1	Control	2539027	26	61	47.5	151.5	37.5	174.5	FALSE
11636.1	Control	2539034	11	26	34.0	16.5	0.0	42.0	TRUE
11636.1	Control	2539039	26	62	54.5	236.5	54.5	263.5	FALSE
11636.1	Control	2539045	25	61	46.0	148.5	46.0	186.0	FALSE
11636.1	Control	2539047	26	61	53.5	275.0	53.5	301.5	FALSE
11636.1	Control	2539050	26	61	53.0	143.0	40.0	144.0	FALSE
11636.1	Control	2539052	26	61	40.5	255.0	40.5	268.0	FALSE
11636.1	Control	2539057	25	61	51.5	174.0	51.5	195.5	FALSE
11636.2	Tamoxifen	2539020	26	61	62.5	65.0	21.0	66.5	FALSE
11636.2	Tamoxifen	2539026	26	61	42.0	67.5	24.5	76.0	FALSE
11636.2	Tamoxifen	2539028	23	53	62.0	50.0	16.0	91.5	FALSE
11636.2	Tamoxifen	2539035	26	61	44.0	27.0	17.5	100.0	FALSE
11636.2	Tamoxifen	2539036	25	60	44.0	40.0	11.5	85.5	FALSE
11636.2	Tamoxifen	2539043	26	61	41.0	10.5	4.0	60.5	FALSE
11636.2	Tamoxifen	2539046	3	5	45.0	57.5	45.0	57.5	TRUE
11636.2	Tamoxifen	2539048	4	6	47.5	71.5	47.5	71.5	FALSE
11636.2	Tamoxifen	2539053	26	62	43.5	41.0	18.5	85.5	FALSE
11636.2	Tamoxifen	2539058	26	61	47.0	52.0	22.5	61.5	FALSE
11636.3	Dexamethasone	2539024	27	61	29.0	147.5	23.0	147.5	FALSE
11636.3	Dexamethasone	2539029	24	57	69.0	117.0	23.0	123.5	FALSE
11636.3	Dexamethasone	2539031	25	61	62.5	220.5	36.5	281.5	FALSE
11636.3	Dexamethasone	2539032	26	61	44.0	269.0	37.5	269.0	FALSE
11636.3	Dexamethasone	2539033	26	61	39.5	174.0	39.5	189.5	FALSE
11636.3	Dexamethasone	2539040	26	61	47.0	135.0	26.0	142.5	FALSE
11636.3	Dexamethasone	2539042	24	57	47.0	219.5	44.0	219.5	FALSE
11636.3	Dexamethasone	2539044	27	61	55.0	309.5	47.5	309.5	FALSE
11636.3	Dexamethasone	2539051	26	61	40.5	323.5	32.5	323.5	FALSE
11636.3	Dexamethasone	2539056	27	61	51.5	280.0	38.5	280.0	FALSE
11636.4	Tamoxifen + Dexamethasone	2539019	3	4	45.5	54.5	45.5	54.5	TRUE
11636.4	Tamoxifen + Dexamethasone	2539022	5	11	41.5	41.5	41.5	68.0	FALSE
11636.4	Tamoxifen + Dexamethasone	2539025	6	9	51.0	60.5	51.0	91.5	FALSE
11636.4	Tamoxifen + Dexamethasone	2539030	26	61	43.0	34.0	13.5	73.5	FALSE
11636.4	Tamoxifen + Dexamethasone	2539037	26	61	43.5	27.5	8.0	54.0	FALSE
11636.4	Tamoxifen + Dexamethasone	2539038	25	58	46.0	27.0	20.5	77.0	FALSE
11636.4	Tamoxifen + Dexamethasone	2539041	26	61	47.0	30.0	13.0	62.0	FALSE
11636.4	Tamoxifen + Dexamethasone	2539049	25	61	32.5	7.5	5.5	32.5	FALSE
11636.4	Tamoxifen + Dexamethasone	2539054	26	61	43.0	33.5	15.0	74.0	FALSE
11636.4	Tamoxifen + Dexamethasone	2539055	25	61	36.0	7.5	3.5	51.0	FALSE

2.4 Tumor growth per mouse

# Plot function for each treatment group separately
plot_group_mice <- function(df_group, title_text) {
  ggplot(df_group, aes(x = day, y = volume_mm3, group = mouse_id, color = mouse_id)) +
    geom_line(linewidth = 1, alpha = 0.9) +
    geom_point(size = 2, alpha = 0.9) +
    scale_color_discrete(guide = "none") +
    labs(
      title = title_text,
      x = "Time (days)",
      y = "Tumor volume (mm³)"
    ) +
    theme_classic(base_size = 12) +
    theme(
      plot.title = element_text(face = "bold"),
      panel.grid.major.y = element_line(color = "grey90", linewidth = 0.25),
      panel.grid.minor.y = element_blank()
    )
}

# Loop over groups and print separate plots
for (g in levels(mouse_df$group_name)) {
  sub <- dplyr::filter(mouse_df, group_name == g)
  n_m <- dplyr::n_distinct(sub$mouse_id)
  ttl <- paste0(g, " — individual tumor growth (n = ", n_m, ")")
  print(plot_group_mice(sub, ttl))
}

2.5 Exclude mice with < 30 days

min_duration_days <- 30

duration_summary <- mouse_df %>%
  group_by(group_code, group_name, mouse_id) %>%
  summarise(
    min_day = min(day, na.rm = TRUE),
    max_day = max(day, na.rm = TRUE),
    duration_days = max_day - min_day,
    .groups = "drop"
  )

exclude_ids <- duration_summary %>%
  filter(duration_days < min_duration_days) %>%
  pull(mouse_id)

mouse_df_filtered <- mouse_df %>%
  filter(!mouse_id %in% exclude_ids)

knitr::kable(
  duration_summary %>%
    mutate(excluded = duration_days < min_duration_days),
  caption = paste("Mice with duration <", min_duration_days, "days are excluded")
)

Mice with duration < 30 days are excluded
group_code	group_name	mouse_id	max_day	duration_days	excluded
11636.1	Control	2539021	8	8	TRUE
11636.1	Control	2539023	61	61	FALSE
11636.1	Control	2539027	61	61	FALSE
11636.1	Control	2539034	26	26	TRUE
11636.1	Control	2539039	62	62	FALSE
11636.1	Control	2539045	61	61	FALSE
11636.1	Control	2539047	61	61	FALSE
11636.1	Control	2539050	61	61	FALSE
11636.1	Control	2539052	61	61	FALSE
11636.1	Control	2539057	61	61	FALSE
11636.2	Tamoxifen	2539020	61	61	FALSE
11636.2	Tamoxifen	2539026	61	61	FALSE
11636.2	Tamoxifen	2539028	53	53	FALSE
11636.2	Tamoxifen	2539035	61	61	FALSE
11636.2	Tamoxifen	2539036	60	60	FALSE
11636.2	Tamoxifen	2539043	61	61	FALSE
11636.2	Tamoxifen	2539046	5	5	TRUE
11636.2	Tamoxifen	2539048	6	6	TRUE
11636.2	Tamoxifen	2539053	62	62	FALSE
11636.2	Tamoxifen	2539058	61	61	FALSE
11636.3	Dexamethasone	2539024	61	61	FALSE
11636.3	Dexamethasone	2539029	57	57	FALSE
11636.3	Dexamethasone	2539031	61	61	FALSE
11636.3	Dexamethasone	2539032	61	61	FALSE
11636.3	Dexamethasone	2539033	61	61	FALSE
11636.3	Dexamethasone	2539040	61	61	FALSE
11636.3	Dexamethasone	2539042	57	57	FALSE
11636.3	Dexamethasone	2539044	61	61	FALSE
11636.3	Dexamethasone	2539051	61	61	FALSE
11636.3	Dexamethasone	2539056	61	61	FALSE
11636.4	Tamoxifen + Dexamethasone	2539019	4	4	TRUE
11636.4	Tamoxifen + Dexamethasone	2539022	11	11	TRUE
11636.4	Tamoxifen + Dexamethasone	2539025	9	9	TRUE
11636.4	Tamoxifen + Dexamethasone	2539030	61	61	FALSE
11636.4	Tamoxifen + Dexamethasone	2539037	61	61	FALSE
11636.4	Tamoxifen + Dexamethasone	2539038	58	58	FALSE
11636.4	Tamoxifen + Dexamethasone	2539041	61	61	FALSE
11636.4	Tamoxifen + Dexamethasone	2539049	61	61	FALSE
11636.4	Tamoxifen + Dexamethasone	2539054	61	61	FALSE
11636.4	Tamoxifen + Dexamethasone	2539055	61	61	FALSE

2.6 Tumor group mean ± SD (Per day, no binning, filtered mice)

tumor_daily <- mouse_df_filtered %>%
  group_by(group_code, group_name, day) %>%
  summarise(
    mean_volume_mm3 = mean(volume_mm3, na.rm = TRUE),
    sd_mm3          = sd(volume_mm3, na.rm = TRUE),
    n_mice          = n(),
    .groups = "drop"
  ) %>%
  arrange(group_name, day)

ggplot(tumor_daily, aes(x = day, y = mean_volume_mm3,
                        color = group_name, group = group_name)) +
  geom_errorbar(aes(ymin = mean_volume_mm3 - sd_mm3,
                    ymax = mean_volume_mm3 + sd_mm3),
                width = 0.3, alpha = 0.9) +
  geom_line(linewidth = 1.0) +
  geom_point(size = 2.6, shape = 16) +
  scale_color_manual(values = pal, breaks = names(pal), name = "Treatment") +
  labs(
    title = "Tumor growth by treatment group (mean ± SD, per day, filtered mice)",
    x = "Day",
    y = "Tumor volume (mm³)"
  ) +
  theme_classic(base_size = 12) +
  theme(
    legend.position = "top",
    legend.title    = element_text(size = 11),
    legend.text     = element_text(size = 10),
    plot.title      = element_text(face = "bold"),
    panel.grid.major.y = element_line(color = "grey90", linewidth = 0.25),
    panel.grid.minor.y = element_blank()
  )

2.7 Tumor group mean ± SD (Binned, filtered mice)

# Custom binning function:
# - keep day 0 separate
# - bin 1–2 as 1, 3–5 as 3, etc.
bin_day_custom <- function(day, width = 3L, anchor = 1L, label = "start") {
  if (day == 0) return(0L)
  label <- match.arg(label, choices = c("start", "center"))
  base <- anchor + width * floor((day - anchor)/width)
  if (label == "center") base <- base + floor((width - 1L)/2L)
  as.integer(base)
}

tumor_binned <- mouse_df_filtered %>%
  mutate(bin_day = vapply(day, bin_day_custom,
                          integer(1), width = bin_width,
                          anchor = bin_anchor,
                          label = bin_label)) %>%
  group_by(group_code, group_name, bin_day) %>%
  summarise(
    mean_volume_mm3 = mean(volume_mm3, na.rm = TRUE),
    sd_mm3          = sd(volume_mm3, na.rm = TRUE),
    n_mice          = n(),
    .groups = "drop"
  ) %>%
  arrange(group_name, bin_day)

ggplot(tumor_binned, aes(x = bin_day, y = mean_volume_mm3,
                         color = group_name, group = group_name)) +
  geom_errorbar(aes(ymin = mean_volume_mm3 - sd_mm3,
                    ymax = mean_volume_mm3 + sd_mm3),
                width = 0.3, alpha = 0.9) +
  geom_line(linewidth = 1.0) +
  geom_point(size = 2.6, shape = 16) +
  scale_color_manual(values = pal, breaks = names(pal), name = "Treatment") +
  labs(
    title = paste0("Tumor growth by treatment group (mean ± SD, ",
                   bin_width, "-day bins, day 0 kept separate)"),
    x = "Day (with bins)",
    y = "Tumor volume (mm³)"
  ) +
  theme_classic(base_size = 12) +
  theme(
    legend.position = "top",
    legend.title    = element_text(size = 11),
    legend.text     = element_text(size = 10),
    plot.title      = element_text(face = "bold"),
    panel.grid.major.y = element_line(color = "grey90", linewidth = 0.25),
    panel.grid.minor.y = element_blank()
  )

3 Bodyweight analysis

3.1 Bodyweight per mouse

4 Bodyweight analysis (unfiltered mice)

4.1 Load bodyweight data

bw_raw <- readxl::read_excel(bw_file, sheet = "Mouse_day_mean")

# Detect correct bodyweight column
bw_candidates <- c("mean_bodyweight_g", "bodyweight_g")
bw_col <- bw_candidates[bw_candidates %in% names(bw_raw)][1]
if (is.na(bw_col)) stop("No bodyweight column found in Mouse_day_mean")

bw_mouse <- bw_raw %>%
  mutate(
    group_code   = as.character(group_code),
    group_name   = factor(unname(group_name_map[group_code]), levels = names(pal)),
    day          = as.numeric(day),
    mouse_id     = as.character(mouse_id),
    bodyweight_g = .data[[bw_col]]
  ) %>%
  filter(group_code %in% names(group_name_map)) %>%
  arrange(group_name, mouse_id, day)

knitr::kable(head(bw_mouse, 8), caption="Preview: bodyweight per mouse per day")

Preview: bodyweight per mouse per day
group_code	group_name	mouse_id	day	bodyweight_g
11636.1	Control	2539021	0	21.9
11636.1	Control	2539021	1	21.6
11636.1	Control	2539021	2	21.3
11636.1	Control	2539021	3	20.6
11636.1	Control	2539021	4	19.8
11636.1	Control	2539021	5	19.1
11636.1	Control	2539021	6	19.8
11636.1	Control	2539021	7	18.5

4.2 Bodyweight per mouse

plot_bw_group <- function(df_group, title_text) {
  ggplot(df_group, aes(x = day, y = bodyweight_g,
                       group = mouse_id, color = mouse_id)) +
    geom_line(linewidth = 1, alpha = 0.9) +
    geom_point(size = 2, alpha = 0.9) +
    scale_color_discrete(guide = "none") +
    labs(
      title = title_text,
      x = "Time (days)",
      y = "Bodyweight (g)"
    ) +
    theme_classic(base_size = 12) +
    theme(
      plot.title = element_text(face = "bold"),
      panel.grid.major.y = element_line(color = "grey90", linewidth = 0.25),
      panel.grid.minor.y = element_blank()
    )
}

for (g in levels(bw_mouse$group_name)) {
  sub <- filter(bw_mouse, group_name == g)
  n_m <- n_distinct(sub$mouse_id)
  ttl <- paste0(g, " — individual bodyweights (n = ", n_m, ")")
  print(plot_bw_group(sub, ttl))
}

4.3 Bodyweight group mean ± SD (per day, unfiltered)

bw_daily <- bw_mouse %>%
  group_by(group_name, day) %>%
  summarise(
    mean_bw = mean(bodyweight_g, na.rm = TRUE),
    sd_bw   = sd(bodyweight_g, na.rm = TRUE),
    n       = n(),
    .groups = "drop"
  ) %>%
  arrange(group_name, day)

ggplot(bw_daily, aes(x = day, y = mean_bw,
                     color = group_name, group = group_name)) +
  geom_errorbar(aes(ymin = mean_bw - sd_bw, ymax = mean_bw + sd_bw),
                width = 0.3, alpha = 0.9) +
  geom_line(linewidth = 1.0) +
  geom_point(size = 2.6, shape = 16) +
  scale_color_manual(values = pal, breaks = names(pal), name = "Treatment") +
  labs(
    title = "Bodyweight by treatment group (mean ± SD, per day, unfiltered)",
    x = "Day",
    y = "Bodyweight (g)"
  ) +
  theme_classic(base_size = 12) +
  theme(
    legend.position = "top",
    legend.title    = element_text(size = 11),
    legend.text     = element_text(size = 10),
    plot.title      = element_text(face = "bold"),
    panel.grid.major.y = element_line(color = "grey90", linewidth = 0.25),
    panel.grid.minor.y = element_blank()
  )

4.4 Bodyweight waterfall: maximale % verlies per muis

# Bereken baseline per muis (dag 0 indien aanwezig, anders eerste meting)
baseline_df <- bw_mouse %>%
  group_by(group_code, mouse_id) %>%
  summarise(baseline_g = {
    d <- arrange(cur_data_all(), day)
    if (any(d$day == 0, na.rm = TRUE)) {
      d$bodyweight_g[d$day == 0][1]
    } else {
      d$bodyweight_g[1]
    }
  }, .groups = "drop")

# % verandering tov baseline
bw_mouse_pct <- bw_mouse %>%
  left_join(baseline_df, by = c("group_code", "mouse_id")) %>%
  mutate(pct_change = 100 * (bodyweight_g - baseline_g) / baseline_g)

# Nadir per muis (laagste % verandering)
nadir <- bw_mouse_pct %>%
  group_by(group_code, group_name, mouse_id) %>%
  summarise(min_pct = min(pct_change, na.rm = TRUE), .groups = "drop")

# Orden muizen per groep (meest negatieve eerst)
wf <- nadir %>%
  group_by(group_name) %>%
  arrange(min_pct, .by_group = TRUE) %>%
  mutate(ix = row_number()) %>%
  ungroup()

# Voeg offsets toe zodat groepen netjes naast elkaar komen
group_sizes <- wf %>%
  count(group_name, name = "n")
gap <- 2
group_sizes <- group_sizes %>%
  mutate(offset = lag(cumsum(n) + gap, default = 0))
wf <- wf %>%
  left_join(group_sizes, by = "group_name") %>%
  mutate(x = ix + offset)

# Posities voor labels op de x-as
centers <- group_sizes %>%
  mutate(center = offset + n/2)

# Plot
ggplot(wf, aes(x = x, y = min_pct, fill = group_name)) +
  geom_col(width = 0.9, alpha = 0.9) +
  geom_hline(yintercept = -10, linetype = "dashed") +
  scale_fill_manual(values = pal, breaks = names(pal), name = "Treatment") +
  scale_x_continuous(breaks = centers$center, labels = centers$group_name) +
  labs(
    title = "Bodyweight waterfall: maximale % verlies per muis",
    x = "Group",
    y = "Maximaal % verlies t.o.v. baseline"
  ) +
  theme_classic(base_size = 12) +
  theme(
    legend.position = "none",
    plot.title = element_text(face = "bold"),
    panel.grid.major.y = element_line(color = "grey90", linewidth = 0.25),
    panel.grid.minor.y = element_blank()
  )

4.5 Tumor: AUC per mouse + ANOVA/T-tests

library(pracma)   # voor trapz()
library(ggpubr)   # voor stat_compare_means()

# 1) Bereken AUC per muis (trapeziumregel, tot laatste dag van die muis)
auc_per_mouse <- mouse_df_filtered %>%
  group_by(group_code, group_name, mouse_id) %>%
  arrange(day, .by_group = TRUE) %>%
  summarise(
    last_day = max(day, na.rm = TRUE),
    auc      = trapz(day, volume_mm3),
    .groups  = "drop"
  )

# Preview van de tabel
knitr::kable(head(auc_per_mouse, 8), caption = "Preview: AUC per mouse")

Preview: AUC per mouse
group_code	group_name	mouse_id	last_day	auc
11636.1	Control	2539023	61	5979.50
11636.1	Control	2539027	61	5872.00
11636.1	Control	2539039	62	8902.75
11636.1	Control	2539045	61	6080.75
11636.1	Control	2539047	61	11832.00
11636.1	Control	2539050	61	5185.75
11636.1	Control	2539052	61	8088.25
11636.1	Control	2539057	61	7786.50

# 2) ANOVA over alle groepen
anova_model <- aov(auc ~ group_name, data = auc_per_mouse)
anova_res <- summary(anova_model)

# 3) Pairwise t-tests met Holm correctie
pairwise_res <- pairwise.t.test(
  auc_per_mouse$auc,
  auc_per_mouse$group_name,
  p.adjust.method = "holm"
)

# 4) Tukey HSD post-hoc
tukey_res <- TukeyHSD(anova_model)
tukey_table <- as.data.frame(tukey_res$group_name)

knitr::kable(
  tukey_table,
  digits = 4,
  caption = "Tukey post-hoc test for pairwise group comparisons (AUC)"
)

Tukey post-hoc test for pairwise group comparisons (AUC)
	diff	lwr	upr	p adj
Tamoxifen-Control	-4990.3750	-7166.210	-2814.540	0.0000
Dexamethasone-Control	-1023.5125	-3087.691	1040.666	0.5392
Tamoxifen + Dexamethasone-Control	-5755.9018	-8008.105	-3503.698	0.0000
Dexamethasone-Tamoxifen	3966.8625	1902.684	6031.041	0.0001
Tamoxifen + Dexamethasone-Tamoxifen	-765.5268	-3017.730	1486.677	0.7912
Tamoxifen + Dexamethasone-Dexamethasone	-4732.3893	-6876.916	-2587.863	0.0000

# 5) Boxplot voor AUC per groep, met p-waardes
ggplot(auc_per_mouse, aes(x = group_name, y = auc, fill = group_name)) +
  geom_boxplot(alpha = 0.7, outlier.shape = NA) +
  geom_jitter(width = 0.2, alpha = 0.7, size = 2) +
  scale_fill_manual(values = pal, guide = "none") +
  labs(
    title = "Tumor burden (AUC) per treatment group",
    x = "Treatment group",
    y = "AUC (mm³ × days)"
  ) +
  theme_classic(base_size = 12) +
  theme(
    axis.text.x = element_text(angle = 20, hjust = 1),
    plot.title  = element_text(face = "bold")
  ) +
  stat_compare_means(
    method = "anova",
    label.y = max(auc_per_mouse$auc) * 1.1
  ) +
  stat_compare_means(
    method = "t.test",
    comparisons = list(c("Tamoxifen", "Tamoxifen + Dexamethasone")),
    label = "p.signif",
    label.y = max(auc_per_mouse$auc) * 1.05
  )

# 6) Samenvattende tabel per groep (mean ± SD)
auc_summary <- auc_per_mouse %>%
  group_by(group_name) %>%
  summarise(
    n     = n(),
    mean  = mean(auc, na.rm = TRUE),
    sd    = sd(auc, na.rm = TRUE),
    .groups = "drop"
  ) %>%
  mutate(mean_sd = paste0(round(mean, 1), " ± ", round(sd, 1)))

knitr::kable(
  auc_summary,
  digits = 1,
  caption = "Summary of AUC per treatment group (mean ± SD, n mice)"
)

Summary of AUC per treatment group (mean ± SD, n mice)
group_name	n	mean	sd	mean_sd
Control	8	7465.9	2187.1	7465.9 ± 2187.1
Tamoxifen	8	2475.6	459.3	2475.6 ± 459.3
Dexamethasone	10	6442.4	2045.6	6442.4 ± 2045.6
Tamoxifen + Dexamethasone	7	1710.0	476.5	1710 ± 476.5

# 7) Print resultaten in de console
anova_res

##             Df    Sum Sq  Mean Sq F value   Pr(>F)    
## group_name   3 193707213 64569071   25.31 3.03e-08 ***
## Residuals   29  73983274  2551147                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

pairwise_res

## 
##  Pairwise comparisons using t tests with pooled SD 
## 
## data:  auc_per_mouse$auc and auc_per_mouse$group_name 
## 
##                           Control Tamoxifen Dexamethasone
## Tamoxifen                 4.0e-06 -         -            
## Dexamethasone             0.37    4.0e-05   -            
## Tamoxifen + Dexamethasone 7.1e-07 0.37      6.2e-06      
## 
## P value adjustment method: holm

Tumor & Bodyweight Analysis with QC and Filtering

Auto-generated Rmd

2025-09-09

1 Setup

2 Tumor analysis

2.1 Shared mapping & helpers

2.2 Tumor: data loading

2.3 Tumor: data quality check

2.4 Tumor growth per mouse

2.5 Exclude mice with < 30 days

2.6 Tumor group mean ± SD (Per day, no binning, filtered mice)

2.7 Tumor group mean ± SD (Binned, filtered mice)

3 Bodyweight analysis

3.1 Bodyweight per mouse

4 Bodyweight analysis (unfiltered mice)

4.1 Load bodyweight data

4.2 Bodyweight per mouse

4.3 Bodyweight group mean ± SD (per day, unfiltered)

4.4 Bodyweight waterfall: maximale % verlies per muis

4.5 Tumor: AUC per mouse + ANOVA/T-tests